Abstract. This paper proposes the combination of the THESEUS multi-criteria sorting method with an evolutionary optimization-based preference-disaggregation analysis. The main features of the combined method are studied by performing an extensive computer experiment that explores many models of preferences and sizes of problems as well as different degrees of decision-maker involvement. As a result of the experiment, the effectiveness of the combined framework and the importance of the decision-maker's involvement are characterized.
To set the parameter values for the outranking model is usually a demanding task for the decision-maker (DM) because it is necessary to provide a large number of parameters (thresholds and weights). The use of indirect methods (preferencedisaggregation analysis (PDA)) to infer the set of parameters from a battery of decision examples allows the DM to avoid the task of specifying these values. In this paper a robustness analysis has been made to a PDA method based on an evolutionary approach. In this case we use THESEUS method as an artificial DM and its assignment rule for evaluating each individual of a multi-objective evolutionary algorithm (MOEA) population. The non-dominated solutions are acceptable parameter settlements.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.